Thoothukudi (lat 8°47ˈN and long 78°9ˈE) is one of the important fish landing centres in south east coast on India [1]. The mesh size of gillnets of Thoothukudi coast range from 28 mm to 128 mm. The fishing season for S. forsteri (Sphyraenidae) in this coast is from November to June. This species form fishery in the coastal region, 20 to 30 nautical miles away from the shore at the depth ranging from 30 m to 40 m.

Selectivity of fishing gear has a direct influence on the exploited stock. Gillnet Selectivity may be estimated by either indirectly or directly. Indirectly, it is inferred by examining the size distribution of gillnet catches or from maximum girth and head girth measurement. By direct method, the selectivity parameters are estimated based on the proportions of fish caught from different size classes of a population with known length frequency distribution [1].

Gillnet selectivity studies in Indian waters are mainly concerned with optimization of mesh size for the commercial exploitation of important marine and fresh water fishes. Attempts have been made by a few researchers’ workers to arrive at optimum length of capture of certain important fish based on the length at first maturity. Knowledge on the length at first maturity of the species to be captures is essential for optimizing the mesh size for sustained commercial exploitation of any species [1].

This species are indiscriminately fished indiscriminately fish and no attempt has so far been made to optimize the mesh size of gillnets to capture S. forsteri. The present study deals with the estimation of the gillnet selection factors of S. forsteri so as to optimize the mesh size for its commercial exploitation.

The study was carried out for one year from April 2012 to March 2013 at Threspuram, Inigonagar and Tharuvaikulam landing centers for traditional crafts of Thoothukudi Coast (Tamil Nadu, Southeast Coast, India). Nets with the mesh size of 28 mm, 50 mm and 52 mm were sampled from the gillnets operated from the commercial fishing boats of Thoothukudi origin, (Popularly called “Vallam”). The nets with the mesh size of 28 mm, 50 mm and 52 mm were named as net ‘A’, ‘B’, and ‘C’ respectively. The features of the net A, B and C are given in Table 1.

Table 1: Description of the gillnets operated for fishing barracuda along Thoothukudi coast.

The webbings were maintained to head rope, maintaining a uniform horizontal hanging co-efficient of 0.5. The total length of individual fish was measured to the nearest cm. The mesh size determined by measuring the stretched meshes with a centimetre scale [2].

The selectivity was estimated by using the indirect method of Sparrae et al. [3]. According to him, for gilling and wedging, the selection curves are bell shaped and can be described by the following relation

Where S (L) is the length based gear selectivity, L is the mid length of the size/length group, Lc is the mean selection length per being caught and S is the standard deviation of the normal distribution.

The procedure for estimation of selection curve involved the following steps:

CB=Number of fish caught in net with larger mesh size (m²)

CA=number if fish caught in A net with smaller mesh size (m¹).

Calculations for log ratios for successive fish lengths Y=Ln (CB/CA)

Linear Regression analysis was made against the mid value of each of the length group as follows

Where, a₁=Intercept for BA combination and b₁=slope value for BA combination.

Where, a₂=Intercept for CB combination, b₂=slope value for CB combination and L=mid length of a length group. CA,CB and CC refers to catch in number in net with the mesh size A,B and C corresponding to the mid value of length group, L.

Regression analysis of the log ratios against the interval midpoint and expressed as,

Where, Y is the natural logarithm of ratio of catches, L is the midpoint of the length class and a and b are constants. The common selection factors (SF) for three mesh combination viz. A, B and C was derived using the equation.

The common standard deviation(s) for three combinations A, B and C was derived using the equation.

The mean selection length (Lc) of each net was estimated from the common selection factor (SF) as follow,

Where, ‘Lci’ refers to mean selection length of different nets ‘mi’ refers to different mesh size.

The optimum mesh size for the commercial exploitation of each species of barracuda were worked out bared on the following formula,

Where ‘m’ is the stretched measure of mesh in cm and ‘Lopt’ is either the mid-length of the commercially significant length group or mean length at first maturity of the respective species in cm. SF is the mean value of selection factors derived based on different mesh combinations.

Monthly Average catch and effect of barracudas are calculated as; the mean daily catch for a month was worked out by taking the mean total catch estimated during different sampling days of months. The average daily catch was multiplied by the number of fishing days of the month for estimated the monthly catch. The effort was expressed in terms of boat days and the catch was expressed in terms of kilograms and also in tonnes.

Fishing season of barracuda was found to be from November to June as reported by Mohammed et al. [4,5]. However, a slight deviation in peak season from that reported by Mohammed et al. [4]. Among the two peaks of fishing season of barracuda in the gillnet fishery, one was found to fall on the months of November and May against the November and March as observed by Mohammed et al. [4]. The reason may be attributed to the gear selected for the study. As 15^th April to 30^th May being closed season for trawl net along the coast of Tamil Nadu, One of the peak seasons has been recorded as March instead of May. The comprehensive study involving catches from all fishing gear such as gillnets, long lines and trawl nets is required to find out the exact fishing of barracuda along Thoothukudi coast regarding depth wise distribution. The peak fishing season found to be from November [4] whose study was based on the catches from commercial trawl nets. The reason may be attributed the fact that the fisherman of Thoothukudi used to switchover to long lining from July to September due to rough weather conditions, though barracudas are reported to migrate and available for fishing in shallow water during south-west monsoon (June to August). Catch and effort particulars of S. forsteri are given in Table 2.

Table 2: Monthly Catch and effort particulars of S. forsteri in gillnet for the period from April 2012- March 2013 in Thoothukudi Coast.

Length frequently distribution of S. forsteri in gillnets of mesh sizes 28 cm, 50 cm and 52 cm is given in Table 3. The total length of S. forsteri caught ranged from 7 cm to 66 cm. The net ‘A’ captures more number of immature and maturing animals than net B and C. So, the net A with 28 mm mesh size is not suitable for capture S. forsteri in Thoothukudi coast.

Table 3: Length frequency distribution and the ratios of natural logarithms of number caught under different length group of S. forsteri in different gillnet combinations.

Selectivity parameters viz, Lb, Lc and Ld increased with the increase in mesh size irrespective of species, indicating clear cut selectivity by the gillnets (Ag). This is an accordance with the findings of Reis et al. [6] who reported the gillnets are rather than species specific. Increase in Lc and SF of Amblygaster sirm with the increase in mesh size have also been documented by many researchers [7-11]. Lb, Lc and Ld values for S. obtustata were given in Table 4 and Figure 1.

Table 4: Selectivity parameters of S. forsteri with respect to different gillnet.

Figure 1: Selection curves of S. forsteri for gillnet types A, B and C.

Selection factor for this species is 7.04 this is accordance with the findings of Andreev [12] who reported that the selection factors generally ranges between 5 and 10.

Optimum mesh size for the S. forsteri is 5.82 cm. It is understood that the net A with the mesh size of 2.8 cm need to be discouraged to capture S. forsteri. Further the mesh size of net B 5 cm was too much closer to the estimated mesh size for S. forsteri (5.8 cm), Therefore the mesh size of net B may be increased to optionally capture S. forsteri. Through fishing with net C with higher mesh size (5.2 cm) was found to better and would considerably reduce the recruitment over fishing of S. forsteri.

Luther et al. [13] has recommended discouraging the use of nets with mesh size less than 28 mm to avoid fishing of juveniles of lesser sardines of Vishakhapatnam Coast. The present investigation also recommended to discouraged the gillnet with 28 mm mesh may lead to growth over fishing S. forsteri in Thoothukudi Coast and hence the use of gillnets with mesh size 28 mm and below should be avoided in Thoothukudi waters to ensure sustained the barracuda fishery.